Just wanted to remove the empty strings that got created:
find . -type f -exec sed -i 's/<< "" <</<</g' '{}' \;
find . -type f -exec sed -i 's/<<"" <</<</g' '{}' \;
I also unstaged tools/schema/tags.cpp, since I don't think I changed
that.
This is the result of running this command in directory src/
find . -type f -exec sed -i 's/\(ERR.*\)\\n\"\;/\1\" << std::endl\;/g' '{}' \;
and carefully inspecting the results.
It is unknown whether the platform still works with the current Wesnoth
code. To ease the SDL 2.0 transition support is removed. If a maintaine
steps up support can be added again.
It is unknown whether the platform still works with the current Wesnoth
code. To ease the SDL 2.0 transition support is removed. If a maintainer
steps up support can be added again.
on visual sudio 2010 errno and EWOULDBLOCK are both defined, but don't work properly. So i changed the code into using WSAGetLastError over errno if available.
Malformed data caused config::error exception to be thrown.
Unfortunately campaignd on catching such an exception had no way to
close the right connection, instead all connections were closed because
the local variable that was supposed to contain connection number wasn't
updated. This is fixed by using boost.exception to attach the necessary
data to config::error objects so it can be used at the catch site.
(part of bug #20205)
The exception in question originated from the read_gz() call and caused
the entire process to abort because of a non-fatal issue occurring in a
secondary thread.
First spotted in campaignd. See the bug report for details.
The #undef was introduced in 2010-10-31T09:00:17Z!guillaume.melquiond@gmail.com, but the commit offers no reason
_why_ the macro was undefined. This patch has been tested by Rhonda on
all Debian architectures and there it caused no build problem, whereas
the 2010-10-31T09:00:17Z!guillaume.melquiond@gmail.com caused several architectures failing to build.
The reason the issue was spotted much later is that before Debian
compiled with the ANA network. It switched to the older network code
while changing the build system from autotools to cmake, so the latter
was thought to be the cause.
Fixes Debian bug #636193 and reverts 2010-10-31T09:00:17Z!guillaume.melquiond@gmail.com.
According to SDL docs, SDLNet_TCP_Send is blocking and if the returned
nubmer of bytes sent is less than expected, it's an error. Previous code
could trigger an infinite loop with the function returning zero and our
loop trying to send again and again.
* made the network code use two separate timeouts, an idle timeout
that is reset each time data is received, and a total wait counter
(that was already there),
* fixed a rather unpleasant milli-/microsecond confusion
